There are countless mechanisms by which the brain learns, with the fundamental unit of learning expressed at the synapse. In vitro studies can track the changing strength of synapses based on principles of coordinated presynaptic and postsynaptic activity. Many of the principles of neural learning have been a part of the animals central nervous system for hundreds of millions of years. At the level of an individual person or animal, these changes are collectively expressed in such a way that it is useful for ones interaction with the environment. For example, upon meeting and getting to know somebody for the first time, one learns not only their basic facial features, but also how they use their face to express themselves. The brain is able to hold onto, and later use, this information and this is undeniably related to the modification of synaptic strengths, probably in the visual system. However, it is very difficult to even begin to understand how this pattern of learning is expressed across different cortical areas. There may be millions or billions of synapses that are affected during the learning of a face. Where are these changes taking place, and how is it that changing one set of synaptic weights will not severely disrupt other previously learned items? These changes must be back compatible. Sometimes new technologies are needed to shed new light on aspects of brain function. In the laboratory, we have recently developed a chronic recording array of inertialess microwires that were capable of following the activity of single neurons across multiple days. A methods paper on this subject was recently published McMahon et al., J Neurophysiol (2014). With this method, we are able to isolate individual neurons to measure each of their action potentials, either spontaneous or in response to visual stimuli. Because of their small size, and the apparent absence of a gliosis reaction, the electrodes are able to routinely maintain isolation of the same neurons for periods that are much longer than have previously been demonstrated in the visual cortex. This new capacity has enormous implications for the types of experiments one can conduct. For example, with this capacity, it means one is able to examine the responses of single neurons longitudinally, not only over a few hours of a session, but also between sessions, and even across weeks and months. In a recently published study, we measured the responses of multiple neurons within fMRI defined face patches McMahon et al, Proc Natl Acad Sci (2014). We presented a large number of static stimuli in order to establish the selectivity of individual neurons a fingerprint of sorts. Then we went back on subsequent days to the same electrodes (which remained permanently implanted in position) and found that the neurons were still there, and that they maintained a virtually identical fingerprint of stimulus responses. In fact, over time periods of months and even exceeding one year, neurons in the recorded area maintained their precise pattern of stimulus response selectivity. This finding is unexpected because it suggest that neurons in a region of the brain ostensibly dedicated to faces do not exhibit updates or adjustments in their activity with natural experience. To examine face patch plasticity further, we asked whether intensive perceptual training with faces might affect neural responses Jones et al, SFN Abstr (2013). In that study, monkeys were trained to report the identity of a set of morphed human faces as well as a set of morphed macaque faces. We tracked that activity of neurons over a period of longer than three months to determine whether this training would have any effect on the response fingerprint. Training commenced after one month. Throughout this period we presented the same basic stimuli and examined the responses for signs of neural plasticity. We were unable to detect any changes associated with learning. This is striking given that we were recording from one of the fMRI face patches, where we demonstrated that neurons responded vigorously and selective to faces. These findings together present the unexpected result that neurons in a face patch do not change in their responses over time. It is important to point out that this result will almost certainly differ across the brain, as it is clear that some areas within the brain must learn about faces. However, this study does provide a new understanding about what is possible at a neural level: once individual neurons in certain areas of the brain acquire a specific set of stimulus-contingent responses, they are able to retain this precise selectivity for much longer than was previously supposed.